Red Hat Bugzilla – Bug 46746
Install fails initializing swap
Last modified: 2007-04-18 12:34:20 EDT
Description of Problem:
During an FTP TUI install of Fairfax beta1, the install stopped with a
fatal error initializing swap on device hda5. A second (seemingly)
identical install with the same options then succeeded.
Only happened once. It's possible that the first attempt did something
with swap that caused the second attempt to succeed.
Steps to Reproduce:
1. Choose "Autopartitioning".
2. Choose "Remove all partitions" with hda [*] checked; I believe that this
machine had two primary ext2 partitions (hda1=256MB and hda2=remaining of
8GB) and nothing else (NO swap) when this install started.
3. The auto-selected configuration looked like this:
/dev/hda1 1-1018 7985M ext3 /
/dev/hda2 1019-1042 188M swap
/dev/hda3 1043-1048 47M ext3 /boot
4. Highlight /dev/hda1 and choose "Delete". (hda2 moves to hda1 etc., free
space listed at end of drive)
5. Add 256M / partition (hda1 moves to hda2 etc., / goes on hda1)
6. Add 512M /var partition (hda1 moves to hda2 etc., /var goes on hda1)
7. Add 2048M /usr partition (hda1 moves to hda2 etc., /usr goes on hda1)
8. Add /export partition with "Fill all available" checked (goes at the end
of the disk) -- when all this is finished we have:
/dev/hda1 1-261 2047M ext3 /usr
/dev/hda2 262-326 509M ext3 /var
/dev/hda3 327-259 258M ext3 /
/dev/hda4 360-1048 5404M Extended
/dev/hda5 360-383 188M swap
/dev/hda6 384-389 47M ext3 /boot
/dev/hda7 390-1048 5164M ext3 /export
[I must say that looks really strange to me... is this "intelligent"
partitioning to a purpose, or is it just the order I did things in?]
9. Fill in a bunch more screens and get the "/tmp/install.log" message.
A dialog titled "Error" that reads: "An error occurred trying to
initialize swap on device hda5. This problem is serious, and the install
cannot continue. Press Enter to reboot your system." F3 shows:
* Detected 96M of memory
* Swap attempt of 96M to 192M
The install should begin.
I think the fact that I began the first attempt with two ext2 partitions
and no swap partition(s) (see above) may be relevant? In any case, I did
the same things the second time, and then it worked.
Jeremy, this looks like more failure to reread partition table bugs.
*** This bug has been marked as a duplicate of 46450 ***